Methods, systems, and apparatus, including computer programs encoded on computer storage media, for scoring candidate routes. One of the methods includes obtaining a predictive model trained on training examples from trip log data, wherein each training example has feature values from a particular trip and a value of a dependent variable that represents an outcome of a portion of the particular trip, wherein the features of each particular trip include values obtained from one or more external data feed sources that specify a value of a sensor measurement at a particular point in time during the trip. Sensor values from one or more external data feed sources of a sensor network are received. Feature values are generated using the sensor values received from the one or more external data feed sources. A predicted score is computed for each route using the feature values for the candidate route.
Legal claims defining the scope of protection, as filed with the USPTO.
3. The computer-implemented method of claim 2, wherein the corresponding waypoint-specific score represents a respective benefit value of arriving at the respective waypoint during a predicted time window indicated by the respective candidate route.
4. The computer-implemented method of claim 2, wherein the plurality of training samples is based on trip log data for previously-completed trips, and wherein the corresponding dependent score comprises, for each respective training waypoint of the particular trip, a waypoint-specific training score that indicates an extent to which a task was successfully performed at the respective training waypoint during a corresponding time window.
5. The computer-implemented method of claim 1, wherein each respective waypoint of the same predetermined plurality of waypoints is associated with a corresponding task to be completed at the respective waypoint, and wherein the corresponding score for the respective candidate route is based on a likelihood of successfully completing the corresponding task of at least one waypoint of the same predetermined plurality of waypoints.
6. The computer-implemented method of claim 1, wherein each respective candidate route of the plurality of candidate routes represents a proposed rerouting for subset of a trip of the vehicle.
9. The computer-implemented method of claim 1, wherein the independent feature values for the respective candidate route are based on sensor data received from one or more remote devices configured to generate one or more of vehicle data, passageway data, weather data, geopolitical data, facility data, or agent data.
10. The computer-implemented method of claim 1, wherein the independent feature values for the respective candidate route represent one or more of agent-specific information, payload-specific information, vehicle-specific information, passageway-specific information, path-specific information, environment-specific information, facility-specific information, or customer-specific information.
11. The computer-implemented method of claim 1, wherein the machine learning model is configured to generate, for each respective path of one or more paths between waypoints in the corresponding ordering of the same predetermined plurality of waypoints specified by the candidate route, a predicted duration for the vehicle to travel the respective path, and wherein the corresponding score is based on the predicted duration for each respective path of the one or more paths.
12. The computer-implemented method of claim 1, wherein the machine learning model is configured to generate, for each respective path of one or more paths between waypoints in the corresponding ordering of the same predetermined plurality of waypoints specified by the candidate route, a predicted distance to be traveled by the vehicle on the respective path, and wherein the corresponding score is based on the predicted distance for each respective path of the one or more paths.
15. The system of claim 14, wherein the corresponding waypoint-specific score represents a respective benefit value of arriving at the respective waypoint during a predicted time window indicated by the respective candidate route.
16. The system of claim 14, wherein the plurality of training samples is based on trip log data for previously-completed trips, and wherein the corresponding dependent score comprises, for each respective training waypoint of the particular trip, a waypoint-specific training score that indicates an extent to which a task was successfully performed at the respective training waypoint during a corresponding time window.
17. The system of claim 13, wherein each respective waypoint of the same predetermined plurality of waypoints is associated with a corresponding task to be completed at the respective waypoint, and wherein the corresponding score for the respective candidate route is based on a likelihood of successfully completing the corresponding task of at least one waypoint of the same predetermined plurality of waypoints.
18. The system of claim 13, wherein the independent feature values for the respective candidate route are based on sensor data received from one or more remote devices configured to generate one or more of vehicle data, passageway data, weather data, geopolitical data, facility data, or agent data.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
June 28, 2023
June 4, 2024
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.